How Music Got Free: The End of an Industry, the Turn of the Century, and the Patient Zero of Piracy

By Stephen Richard Witt

12 min read

Why does it matter? Because the piracy revolution wasn't a wave — it was a handful of people in specific rooms making specific choices

Most of the music that defined the early 2000s — the albums that leaked weeks before release, the tracks that flooded dorm rooms before anyone had paid a cent — didn't come from millions of anonymous strangers acting in loose parallel. It came from one man in Shelby, North Carolina, working a factory shift, hiding a disc behind a belt buckle. The digital revolution looks, from a distance, like a weather system: vast, impersonal, no individual fingerprints. Up close, it's a series of specific people making specific decisions, most of them invisible to the industry they were quietly destroying. A German engineer obsessing over a single folk song in a listening lab. A music executive who understood hits but couldn't recognize a technology threat if it walked up and served him papers. And a factory worker who beat a billion-dollar security apparatus with a rubber glove. This is the story of how they all collided — and why only one of them saw it coming, even as he spent years insisting it shouldn't.

The Industry Counted Its Money While the Future Was Being Built in a German Listening Lab

In the spring of 1995, a conference room in Erlangen, Germany hosted the quiet death of a technology. Karlheinz Brandenburg, a lanky, hunching engineer with wire-frame glasses and an unfortunate beard, distributed a fifty-page technical rebuttal to a committee that had already made up its mind. Brandenburg gathered his team afterward and explained, with a forced smile, that the people in charge had simply made a mistake. Again.

This was the seventh time.

The losses weren't about audio quality — Brandenburg's team kept winning the listening tests. The problem was that they'd been talked into wearing an ugly outfit to every competition, by the same people judging the contest.

What Brandenburg and his co-inventor Bernhard Grill had actually built was remarkable. They started with Suzanne Vega's 'Tom's Diner' — an a cappella track whose unaccompanied voice exposed every flaw the algorithm couldn't hide — and spent years running it through thousand-dollar Japanese headphones alongside castanets, jet engines, and hockey crowds, isolating glitches until the encoder was clean. The underlying insight was anatomical: the ear masks quieter sounds near louder ones, suppresses certain frequencies, fills in brief gaps before a loud noise hits. Discard what listeners can't hear anyway, and a full CD's 1.4 million bits per second collapses to 128,000 without the difference being audible.

By the time the committees finished voting, none of it had mattered officially. But Grill had quietly posted a free encoder on the internet. Someone cracked the player software. The files started moving. And the format war the standards bodies declared over was resolved instead by teenagers in dormitories — including a Stanford freshman named David Weekly, who would shortly do something with those files that made the record industry's lawyers very unhappy — who never attended a single meeting.

The Music Industry's Blindness Wasn't Bad Luck — It Was a Document

The industry's failure to see the mp3 coming wasn't a tragedy of circumstance. It was a documented institutional choice, filed with the Securities and Exchange Commission in November 1998, available to anyone who asked.

The most current technology risk the Seagram team could name was a sixteen-year-old Alan Greenspan study on how the dual-head cassette deck had hurt music sales in the early 1980s. They'd paid ten billion dollars to acquire PolyGram from Philips, and securities law required them to catalogue every material risk. They took the obligation seriously. They discussed physical bootlegging of compact discs — serious enough that PolyGram had already abandoned entire markets in Asia and Latin America to organized counterfeiting operations. They wrote about economic recession, about losing bidding wars for artists, about consumer tastes shifting. They listed what they called 'key man' risk: the chance that Doug Morris might have a stroke or be hit by falling debris from space.

The prospectus did not mention the internet. It did not mention broadband, or personal computers, or audio compression. It did not mention file-sharing. It did not mention the mp3.

This was 1998. Winamp had already been downloaded fifteen million times. An eighteen-year-old Stanford freshman named David Weekly had put 110 music files on a university server — in the same months the prospectus was being drafted — and was consuming 80 percent of Stanford's outgoing network bandwidth, with two thousand visitors a day. The Warez Scene had been bundling free mp3 encoders and cracked players in chat rooms for two years. Fraunhofer's own FTP servers were being hammered by traffic from stolen serial numbers. The prospectus writers knew about home taping in 1982; they missed the thing actually happening outside their windows in 1998.

A year earlier, Brandenburg had flown to Washington and shown the RIAA a copy-protected mp3 — a legal distribution mechanism that could have channeled the format's popularity into revenue. The RIAA passed, partly because licensing the codec would have been expensive. Brandenburg flew home. The industry told itself this was a quality problem. What it was, was a customer problem: consumers who had survived scratched vinyl and tinny AM transistor radios were not waiting for perfect sound. They were waiting for convenient sound.

The Seagram prospectus is the industry's blindness made physical. Not a metaphor. A document.

The Patient Zero of Internet Piracy Was a Factory Worker Who Figured Out Belt Buckles

Dell Glover had been watching the guards for months. Every shift, he shuffled toward the exit behind a line of coworkers, waiting to see who got the red light. The white guys wore big oval belt buckles with the Confederate flag painted on. The Black guys wore gilt plates spelling out BOSS in fake diamonds. The Hispanic guys wore longhorn skulls with ornate Western trim. Every single buckle set off the wand. Every single time, the guard shrugged and waved them through.

Glover was born in 1974 into a line of Black tinkerers in Shelby, North Carolina — a grandfather who repaired televisions on the side, a father who fixed engines, men who carved out technical niches during Jim Crow because that was the exit available to them. Dell inherited the disposition. At fifteen, he bought a $2,300 computer from Sears by working weeknights at Shoney's until eleven, commuting straight from school, eventually running the kitchen. He was the kind of person who takes machines apart to see how they work, who runs the mental accounting without a ledger, who notices things other people look at without seeing.

By 2000 he was making $11 an hour at the Universal pressing plant in Kings Mountain — sometimes $16.50 on overtime, sometimes seventy hours a week — and the plant was producing half a million CDs a day. The security apparatus was genuinely impressive: randomized wand searches, X-ray conveyors for bags, alarmed emergency exits, no personal electronics, digitally logged pressing machines, robotic pallets, laser-guided warehouse vehicles, bar-code inventory tracking down to the individual disc. But Glover, standing at the packaging line in his surgical gloves, had found the one gap nothing digital could close: the plastic grinder. When ornate packaging introduced overstock discs that needed to be destroyed at shift's end, the grinder crushed them to shards. No memory. No records. Entirely outside the inventory system. If he was told to destroy 24 discs and only 23 went in, no one in accounting would ever know.

So: wrap an overstock disc in a surgical glove, tuck it into the grinder's control panel during the shift, retrieve it at day's end, wedge it against your waistband. Walk toward the turnstile with an oversized belt buckle positioned directly in front of the disc. When the wand whines, stand still. The guard will shrug. He always does.

He called it security theater, which is exactly right: a pantomime designed to deter, not to catch. The belt buckle that beat the system wasn't an exotic piece of equipment. Everybody in small-town North Carolina already owned one.

Napster Didn't Kill the Music Industry — Winning the Wrong Lawsuit Did

The music industry's most consequential decision in the digital era was a courtroom win.

The day after the 2000 Grammys, RIAA chief Hilary Rosen gathered a few dozen label executives at a Beverly Hills hotel and demonstrated what Napster actually was. Staffers pulled up the service and let the room call out requests — not current hits but deep album cuts, obscure releases, things with no commercial profile whatsoever. The service found every one, instantly. Then someone requested an *NSYNC song that had been on radio for exactly three days and wasn't available for purchase anywhere. It was there. The room went quiet. A Sony executive, trying to release the tension, asked whether suing the company would even be enough.

It was a better question than he knew.

Napster was shut down in July 2001. What nobody had gamed out was what the shutdown actually produced: hundreds of millions of mp3 files sitting on home computers across the country, legally stranded, with no easy way to get more. The format had already escaped into the wild — the lawsuit couldn't recall it, only cut off the supply line. And here is the structural trap the industry walked into: mp3s had been boosting CD sales precisely because, without a reliable portable player, they were an inferior good. You couldn't take them anywhere. The mp3 was interesting; the compact disc was still useful. The moment a well-designed pocket player arrived, that calculation would invert permanently.

Which brings in the other lawsuit. The RIAA had also sued Diamond Multimedia over its Rio mp3 player and lost. That ruling meant portable players could be manufactured and sold without legal interference. Shutdown created demand without supply; the Diamond decision kept the supply vehicle legal. Apple just had to build it.

The iPod launched in October 2001, three months after Napster's servers went dark. iTunes followed in 2003. The industry had spent its legal energy killing the distribution network for free files and left the device market wide open. The $14 compact disc, already undermined, now had a clean replacement waiting. The industry won the lawsuit it needed to lose.

The Piracy Underground Had Rules, Rankings, and a Competitive Culture More Organized Than Most Corporations

At its peak, Rabid Neurosis had fewer than fifty IP addresses authorized to access its private chat channel. The entire operation that flooded the internet with prerelease music — four of the top five bestselling albums in America in 2005, released weeks before anyone could buy them — ran through a network of people who could fit in a school gymnasium. What felt like a diffuse cloud of chaos was actually a small number of people in specific rooms making specific choices according to written rules.

The rules existed on paper. A 5,000-word standards document, drafted by a council of elite piracy groups styling themselves 'the other RIAA,' specified exactly how an mp3 had to be encoded, named, and catalogued before it could be officially released into the Scene. Bit rates, naming conventions, anti-duplication protocols — the document was more rigorous than most corporate style guides.

Kali, the leader of RNS, ran his operation the way a serious analyst runs a portfolio. He read Billboard like a racing handicapper reads a form sheet, tracing corporate acquisitions and pressing agreements to determine which album would be manufactured at which plant on which date. The groups enforcing these standards kept league tables. They tracked who scooped whom by how many days, competing for first-release credit with the intensity of financial traders competing for a tick's advantage on a price move. When Kali identified a target, he directed his sources accordingly. Simon Tai cultivated a DJ slot at Penn's campus radio station for two years specifically to get unmonitored access to promotional discs. The British members leveraged journalism credentials. The Japanese members covered the early-release Tokyo market. Dell Glover, as ADEG, occupied the most valuable position of all: inside the manufacturing line itself. None of them knew each other's real names. Most didn't know each other existed.

The competitive logic was relentless. Leak three weeks early and investigators could narrow the source to a single plant. Leak one week early and a rival group might scoop you anyway. Two weeks was the sweet spot, and Kali held that interval like a conductor keeping time. The legitimate record business, for all its corporate resources, never developed anything as precisely calibrated as the system a twenty-something in the 818 area code was running from the San Fernando Valley.

Glover's Accidental Experiment Produced the Cleanest Data Point in the Whole Piracy Debate

What if the conventional wisdom about music piracy — that leaking an album early destroyed its sales — was simply wrong? Glover's final act before his arrest accidentally produced an answer, and it came from the last place anyone expected: a personal grudge.

The Kanye West versus 50 Cent release-date rivalry in September 2007 was real theater, engineered by Doug Morris at Vivendi Universal — both artists distributed by the same corporate parent, their public beef a marketing device dressed up as old-school hip-hop competition. Kali recognized the stunt immediately and wanted both albums as a matter of principle. Glover held the power to deliver them. He also, after years of taking orders, held something else: a reason to route one album around Kali entirely.

Glover had recently connected with a rival group called Old Skool Classics. When both albums arrived at the pressing plant in mid-August, weeks ahead of their official September 11 release, he made a quiet decision. He sent Kanye's Graduation to OSC and held Curtis for Kali, staggered four days apart. He was finally done being managed.

Graduation hit the piracy networks on August 30. Curtis followed on September 4. When the official release date arrived and SoundScan tallied the first week, Graduation had sold close to a million copies; Curtis moved around 600,000. The album leaked first — and leaked earlier, and to a competing group — outsold the other by nearly 400,000 copies. Glover had expected the opposite. The industry's entire lawsuit strategy rested on the assumption that early leaks suppressed sales. His inadvertent experiment pointed somewhere else entirely: the leak might have been functioning as promotion, seeding demand rather than cannibalizing it. It was a cleaner test than anything the RIAA ever commissioned, and the man who ran it was a factory worker on overtime pay, acting out of spite.

The Man Who Couldn't Recognize a Good Technology Person Made Hundreds of Millions From a YouTube Visit

Morris is sitting in his teenage grandson's bedroom — this titan of the music business, responsible for Eminem and Fifty Cent and Jay-Z, who has just survived a decade of catastrophic technological disruption — watching the 'In Da Club' video on YouTube. He loved the song because it had moved eight million units. His grandson loved it because it bumped. Neither of those reasons is the point. The point is the small rectangular advertisement flashing in the sidebar. Weight-loss supplements. Mortgage refinancing. Some mom with a belly tip. Morris stared at those ads and understood, perhaps for the first time, that someone was making money from his product, and it wasn't him.

The next morning he called his deputy Zach Horowitz. They're selling ads against our videos, Morris said. Horowitz explained, patiently, that the videos were promotional. Morris asked what exactly they were promoting — Get Rich or Die Tryin' had come out four years ago. Shortly after, thousands of videos disappeared from YouTube, and Universal's entire catalog went dark across every major hosting site. The pulldown swept up everything scored with Universal music: wedding tribute videos, fan montages, cage-fighting highlight reels. The outcry was enormous. Morris didn't care. The sites eventually came back with licensing checks. He had created hundreds of millions in revenue from an afternoon visit with his grandson.

A year earlier, in a now-infamous interview with Wired, Morris had explained that he wouldn't be able to recognize a good technology person — that anyone with a plausible story could have deceived him. The internet savaged him for it. And the critics weren't wrong about the diagnosis: Morris had spent the better part of a decade unable to translate his extraordinary instinct for market signals into anything digital. He'd watched Vivendi burn tens of billions on the internet while he stayed profitable by simply asking for less money each year.

But the YouTube visit suggests something stranger than the 'dinosaur' story: the same market reflex that made him great in 1985 — watch the customer, follow the money, move — eventually found its footing in 2007. The man who couldn't recognize a good technology person recognized the ad box.

The Most Sophisticated Piracy Operation in History Was Defeated by Middle-Aged Texans Who Thought the Punishment Was Too Harsh

Three sheriff's deputies were waiting by Glover's truck in the pre-dawn dark of the EDC parking lot. He walked toward them across the asphalt, pulled the key fob from his pocket, and pressed it. The truck chirped. The deputies drew their guns. They told him to raise his hands — he did, the fob still clutched in his fingers — and explained that FBI agents were currently inside his house.

At home he found bulletproof vests and a SWAT team on the lawn, a kicked-in front door, his girlfriend Karen in the kitchen holding their infant son, crying. Special Agent Peter Vu introduced himself and told Glover he'd been looking for him for more than five years. Dockery had already talked. Glover read the search warrant carefully, hoping its terms didn't cover his truck's CD player, which contained a leaked copy of Kanye's Graduation.

That same day, before Glover had finished talking to the FBI, Kali called. Vu had told Glover to play dumb and draw him in. Instead, Glover warned him off: they hit me yesterday, shut it down. Kali said he appreciated it and hung up. That act of loyalty — protecting the man who had controlled and manipulated him for years — would cost Glover everything.

The FBI traced Kali through subpoenas following login credentials on Glover's seized computer to a Time Warner IP registered to Bilkish Cassim in Granada Hills. Her son Adil, 29, was an IT administrator living in the house. On a burned disc in his bedroom, his résumé's document properties listed the author as 'Kali.' That was enough.

The trial in Houston was decided by a jury from which every juror under 40 had been struck, along with anyone who had ever downloaded music legally. The resulting panel — middle-aged Texans firmly in the compact disc era — sat in a fluorescent courtroom for five days and heard the government's case against the men who had leaked 20,000 albums. Then they acquitted both Cassim and his co-defendant. Jurors said afterward they understood the defendants were probably guilty. They simply thought the potential punishment was disproportionate to what the men had actually done, so they voted no. Jury nullification: the legal system's quiet admission that a law can be technically valid and socially unconvincing at the same time.

Glover, who had pleaded guilty and cooperated expecting leniency, went to prison for three months. The man he'd warned walked free. The most sophisticated music piracy operation in history — 20,000 albums, eleven years, a global network of topsites — ended not with a decisive reckoning but with a Texas jury quietly declining to participate. Whether that's an injustice or a verdict on the laws themselves is the question the book sets down without answering.

The Belt Buckle and the Nail Gun

Here is what the book leaves you with: Dell Glover in Shelby, fixing a neighbor's router for twenty dollars, glancing out from behind the curtain at a street that looks the same as it did before he changed the world. Stephen Witt — the journalist who spent years collecting the files — in a Queens warehouse, listening to a nail gun fire six times through each of his hard drives: a hundred thousand songs reduced to a rattle of broken cores, then dropped into a dumpster already full of someone else's. Neither man is triumphant. Neither image resolves. Was any of this avoidable, and if so, by whom? The engineers, the executives, the factory worker, the jury — everyone had a moment when a different choice was possible. Nobody took it. The revolution happened anyway, and when it was over, the people who built it dismantled the evidence by hand.